mloss.org new softwarehttp://mloss.orgUpdates and additions to mloss.orgenThu, 14 Dec 2017 02:39:19 -0000Aboleth 0.7http://mloss.org/revision/view/2139/<html><p>A bare-bones TensorFlow framework for Bayesian deep learning and Gaussian process approximation with stochastic gradient variational Bayes inference.
</p>
<p>Some of the features of Aboleth:
</p>
<ul>
<li><p>Bayesian fully-connected, embedding and convolutional layers using SGVB for inference.
</p>
</li>
<li><p>Random Fourier and arc-cosine features for approximate Gaussian processes. Optional variational optimisation of these feature weights.
</p>
</li>
<li><p>Imputation layers with parameters that are learned as part of a model.
</p>
</li>
<li><p>Very flexible construction of networks, e.g. multiple inputs, ResNets etc.
</p>
</li>
<li><p>Optional maximum-likelihood type II inference for model parameters such as weight priors/regularizers and regression observation noise.
</p>
</li>
<li><p>Compatible and interoperable with other neural net frameworks such as Keras (see the demos for more information).
</p>
</li>
</ul></html>daniel steinberg, lachlan mccalman, louis tiao, , simon ocallaghan, alistair reidThu, 14 Dec 2017 02:39:19 -0000http://mloss.org/software/rss/comments/2139http://mloss.org/revision/view/2139/deep learningvariational inferencegaussian processtensorflowsparkcrowd 0.1.5http://mloss.org/revision/view/2138/<html><p>The use of crowdsourcing for labelling data for machine learning introduces several complications: the annotators may not understand the problem correctly, they may not have the expertise required, they may be random annotators or even try to deteriorate the results. To learn from this labels in contexts of Big Data, practitioners need to take into consideration, in some way, the quality of the annotators labelling the data, as these is crucial when the annotations are scarce.
This package implements several methods for dealing with this situations using Apache Spark, to facilitate the transition to big scale problems.
</p></html>Enrique G. Rodrigo, Juan A. Aledo, Jose A. GamezWed, 13 Dec 2017 13:13:35 -0000http://mloss.org/software/rss/comments/2138http://mloss.org/revision/view/2138/distributedmachine learningcrowdsourcingsparkAika 0.12http://mloss.org/revision/view/2136/<html><p>Aika is a Java library that automatically extracts and annotates semantic information into text. In case this information is ambiguous, Aika will generate several hypothetical interpretations concerning the meaning of the text and pick the most likely one. The Aika algorithm is based on various ideas and approaches from the field of AI such as artificial neural networks, frequent pattern mining and logic based expert systems. It can be applied to a broad spectrum of text analysis task and combines these concepts in a single algorithm.
</p>
<p>Aika allows to model linguistic concepts like words, word meanings (entities), categories (e.g. person name, city), grammatical word types and so on as neurons in a neural network. By choosing appropriate synapse weights, these neurons can take on different functions within the network. For instance neurons whose synapse weights are chosen to mimic a logical AND can be used to match an exact phrase. On the other hand neurons with an OR characteristic can be used to connect a large list of word entity neurons to determine a category like 'city' or 'profession'.
</p>
<p>Aika is based on non-monotonic logic, meaning that it first draws tentative conclusions only. In other words, Aika is able to generate multiple mutually exclusive interpretations of a word, phrase, or sentence, and select the most likely interpretation. For example a neuron representing a specific meaning of a given word can be linked through a negatively weighted synapse to a neuron representing an alternative meaning of this word. In this case these neurons will exclude each other. These synapses might even be cyclic. Aika can resolve such recurrent feedback links by making tentative assumptions and starting a search for the highest ranking interpretation.
</p>
<p>In contrast to conventional neural networks, Aika propagates activations objects through its network, not just activation values. These activation objects refer to a text segment and an interpretation.
</p>
<p>Aika consists of two layers. The neural layer, containing all the neurons and continuously weighted synapses and underneath that the discrete logic layer, containing a boolean representation of all the neurons. The logic layer uses a frequent pattern lattice to efficiently store the individual logic nodes. This architecture allows Aika to process extremely large networks since only neurons that are activated by a logic node need to compute their weighted sum and their activation value. This means that the fast majority of neurons stays inactive during the processing of a given text.
</p>
<p>To prevent that the whole network needs to stay in memory during processing, Aika uses the provider pattern to suspend individual neurons or logic nodes to an external storage like a mongo db.
</p></html>Lukas MolzbergerTue, 12 Dec 2017 15:21:11 -0000http://mloss.org/software/rss/comments/2136http://mloss.org/revision/view/2136/information extractioninferenceneural networktext miningTheano 1.0.1http://mloss.org/revision/view/2135/<html><p>Theano is a Python library that allows you to define, optimize, and evaluate mathematical expressions involving multi-dimensional arrays efficiently. Theano features:
</p>
<pre><code>* tight integration with numpy – Use numpy.ndarray in Theano-compiled functions.
* transparent use of a GPU – perform data-intensive computations much faster than on a CPU.
* symbolic differentiation – Let Theano do your derivatives.
* speed and stability optimizations – Get the right answer for log(1+x) even when x is really tiny.
* dynamic C code generation – Evaluate expressions faster.
* extensive unit-testing and self-verification – Detect and diagnose many types of mistake.
</code></pre><p>Theano has been powering large-scale computationally intensive scientific investigations since 2007. But it is also approachable enough to be used in the classroom (IFT6266 at the University of Montreal).
</p>
<p>Theano has been used primarily to implement large-scale deep learning algorithms. To see how, see the Deep Learning Tutorials (http://www.deeplearning.net/tutorial/)
</p></html>mostly LISA labThu, 07 Dec 2017 14:14:38 -0000http://mloss.org/software/rss/comments/2135http://mloss.org/revision/view/2135/pythoncudagpusymbolic differentiationnumpyr-cran-CoxBoost 1.4http://mloss.org/revision/view/1313/<html><p>Cox models by likelihood based boosting for a single survival endpoint or competing risks: This package provides routines for fitting Cox models by likelihood based boosting for a single endpoint or in presence of competing risks
</p></html>Harald BinderFri, 01 Dec 2017 00:00:05 -0000http://mloss.org/software/rss/comments/1313http://mloss.org/revision/view/1313/r-cranr-cran-e1071 1.6-8http://mloss.org/revision/view/2061/<html><p>Misc Functions of the Department of Statistics, Probability Theory Group (Formerly: E1071), TU Wien: Functions for latent class analysis, short time Fourier transform, fuzzy clustering, support vector machines, shortest path computation, bagged clustering, naive Bayes classifier, ...
</p></html>David Meyer [aut, cre], Evgenia Dimitriadou [aut, cph], Kurt Hornik [aut], Andreas Weingessel [aut], Friedrich Leisch [aut], Chih-Chung Chang [ctb, cph] (libsvm C++-code), Chih-Chen Lin [ctb, cph] (liFri, 01 Dec 2017 00:00:05 -0000http://mloss.org/software/rss/comments/2061http://mloss.org/revision/view/2061/r-cranr-cran-caret 6.0-77http://mloss.org/revision/view/2120/<html><p>Classification and Regression Training: Misc functions for training and plotting classification and regression models.
</p></html>Max Kun, Jed Wing, Steve Weston, Andre Williams, Chris Keefer, Allan Engelhardt, Tony Cooper, Zachary Mayer, Brenton Kenkel, the R Core Team, Michael Benesty, Reynald Lescarbeau, Andrew Ziem, Luca ScrFri, 01 Dec 2017 00:00:05 -0000http://mloss.org/software/rss/comments/2120http://mloss.org/revision/view/2120/r-cranr-cran-Boruta 5.2.0http://mloss.org/revision/view/2053/<html><p>Wrapper Algorithm for All Relevant Feature Selection: An all relevant feature selection wrapper algorithm. It finds relevant features by comparing original attributes' importance with importance achievable at random, estimated using their permuted copies.
</p></html>Miron Bartosz Kursa [aut, cre], Witold Remigiusz Rudnicki [aut]Fri, 01 Dec 2017 00:00:04 -0000http://mloss.org/software/rss/comments/2053http://mloss.org/revision/view/2053/r-cranGPML Gaussian Processes for Machine Learning Toolbox 4.1http://mloss.org/revision/view/2134/<html><p>The GPML toolbox implements approximate inference algorithms for Gaussian processes such as Expectation Propagation, the Laplace Approximation and Variational Bayes for a wide class of likelihood functions for both regression and classification. It comes with a big algebra of covariance, likelihood, mean and hyperprior functions allowing for flexible modeling. The code is fully compatible to Octave 3.2.x.
</p></html>Carl Edward Rasmussen, Hannes NickischMon, 27 Nov 2017 19:26:13 -0000http://mloss.org/software/rss/comments/2134http://mloss.org/revision/view/2134/classificationregressionapproximate inferencegaussian processesDFLsklearn, Hyperparameters optimization in Scikit Learn 0.1http://mloss.org/revision/view/2133/<html><p>DFLsklearn is a method that performs cross validation over the hyperparameters of the Scikit-learn methods based on an efficient derivative free mixed-integer line search algorithm called Derivative Free Line-search (DFL). DFL is an algorithm with deterministic convergence properties toward local stationary points of the objective function. Furthermore, the DFL algorithm is implemented in a highly optimized Fortran code.
</p>
<p>This software is focused on performing the hyperparameters optimization for each single estimator of Scikit-learn, enabling expert users to exploit as much as possible the features of the machine learning method they are using.
</p>
<p>The source code of DFLsklearn is available here on jmlr.org and GitHub (url{https://github.com/midagroup/DFLsklearn}) under the New BSD License. The code follows PEP8 standards.
A setup.py file is provided in order to easily compile the Fortran code and install the module in the PythonPath.
DFLsklearn just depends on Scikit-Learn package for easy portability and compatibility on different platforms.
</p></html>Vittorio Latorre, Federico BenvenutoThu, 23 Nov 2017 13:14:36 -0000http://mloss.org/software/rss/comments/2133http://mloss.org/revision/view/2133/machine learninghyperparameter selection